See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/318907542

Language-based personality: A new approach to personality in a digital world

Article  in  Current Opinion in Behavioral Sciences · August 2017

DOI: 10.1016/j.cobeha.2017.07.017

CITATIONS READS

12 1,064

2 authors:

Ryan L. Boyd James W. Pennebaker

University of Texas at Austin University of Texas at Austin
51 PUBLICATIONS   602 CITATIONS    326 PUBLICATIONS   36,145 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Measurement and psychosocial factors of well being View project

Collective memory and social representations of history View project

All content following this page was uploaded by Ryan L. Boyd on 03 October 2017.

The user has requested enhancement of the downloaded file.

 Available online at www.sciencedirect.com
ScienceDirect
Language-based personality: a new approach to
personality in a digital world$
Ryan L Boyd and James W Pennebaker
Personality is typically defined as the consistent set of traits,
attitudes, emotions, and behaviors that people have. For
several decades, a majority of researchers have tacitly agreed
that the gold standard for measuring personality was with
self-report questionnaires. Surveys are fast, inexpensive, and
display beautiful psychometric properties. A considerable
problem with this method, however, is that self-reports reflect
only one aspect of personality — people’s explicit theories of
what they think they are like. We propose a complementary
model that draws on a big data solution: the analysis of the
words people use. Language use is relatively reliable over time,
internally consistent, and differs considerably between people.
Language-based measures of personality can be useful for
capturing/modeling lower-level personality processes that are
more closely associated with important objective behavioral
outcomes than traditional personality measures. Additionally,
the increasing availability of language data and advances in
both statistical methods and technological power are rapidly
creating new opportunities for the study of personality at ‘big
data’ scale. Such opportunities allow researchers to not only
better understand the fundamental nature of personality, but at
a scale never before imagined in psychological research.
Address
The University of Texas at Austin, Department of Psychology,
United States
Corresponding authors: Boyd, Ryan L (ryanboyd@utexas.edu),
Pennebaker, James W (pennebaker@utexas.edu)
Current Opinion in Behavioral Sciences 2017, 18:63 68
This review comes from a themed issue on Big data in the behavioral
sciences
Edited by Michal Kosinski and Tara Behrend
http://dx.doi.org/10.1016/j.cobeha.2017.07.017
2352 1546/ã 2017 Elsevier Ltd. All rights reserved.
People differ dramatically in the ways they think, feel,
and behave in general, forming the basis for what we refer
to as personality. Going back to the ancient Greeks,
$
Preparation of this manuscript was aided by grants from the National
Institutes of Health [5R01GM112697 02], John Templeton Foundation
[#48503], and the National Science Foundation [IIS 1344257].
www.sciencedirect.com
formal thinking about personality has relied on different
methods to measure and explain personality. Classically,
Galen posited four general temperaments sanguine,
phlegmatic, melancholic, and choleric based on his
observations of biology and the theories of Hippocrates
[1]. Freud [2] revolutionized the broader discussion about
personality by arguing that inborn temperament and early
experiences shaped what people were like later in life.
Temperament researchers focused on the activity levels
and emotionality of infants to posit the likely genetic and
biological bases of individual differences [3]. Others, such
as Gordon Allport [4] pointed to the enduring and stable
behavioral styles that people possessed including
the ways they walked, gestured, or chewed gum. Even
the most nuanced behaviors revealed people’s basic
characteristics.
Not until the advent of modern social science did psy-
chologists begin to focus on the careful measurement of
personality [5 7]. In the last quarter of the 20th century,
the trait approach emerged that effectively defined mod-
ern personality theory, ushering in detailed factor models
of the construct [8,9]. The new trait approach energized
the field of personality research, in part because it leaned
heavily on self-reports of participants’ self-concepts for
understanding their general personality characteristics.
This was a profound development in personality research:
widespread adoption of self-reports meant that it was now
possible to have very large groups of people complete
extensive personality scales rather than relying on more
time-intensive and resource-intensive approaches. Paired
with advances in statistical and other computational
methods, the adoption of self-report scales resulted in
new ways of studying the domains and correlates of traits.
Self-report questionnaires can provide rich information
about peoples’ conscious, explicit self-concepts. How-
ever, most personality experts have harbored occasional
doubts about the degree to which people’s self-reported
traits reflect who they really are [10]. For example, to
what degree do self-theories map onto their actual beha-
viors? Across thousands of studies, we know that self-
reports correlate nicely with other self-reports from the
same people, yet often show lackluster overlap with
more objective measures that presumably capture the
same underlying traits. Researchers consistently find that
widely-used and well-validated self-report measures are
insufficient when it comes to forming an accurate under-
standing of even basic human patterns such as workplace
behaviors [11], physical activity [12], and expressions of
happiness [13] or other emotional states [14].
Current Opinion in Behavioral Sciences 2017, 18:63 68
64 Big data in the behavioral sciences
Are we thinking about personality in the right way? Are
people’s self-theories the appropriate gold standard for
assessing personality? If not self-reports, does a gold
standard exist? As we outline below, we must move
beyond the gold standard way of thinking. Self-reports
reflect one dimension of personality, while nervous sys-
tem activity may serve as another, genetic factors may be
the basis of a third, and so on.
Beyond self-reports and biological markers, recent
research has demonstrated that a powerful reflection of
personality can be gleaned from the words people use in
everyday life. As an increasing number of studies dem-
onstrate, the ways in which people use words is reliable
over time, internally consistent, predictive of a wide range
of behaviors and even biological activity, and varies
considerably from person to person. Language, then, is
yet another fundamental dimension of personality. Of
great benefit to researchers, and unlike other standard
personality markers, people do not need to complete
questionnaires or submit to invasive blood or genetic
tests in order to provide useful personality data in the
form of language.
Language and personality in the land of big
data
Over half of the planet’s population uses the internet, and
over 80% of people in developed countries are internet
users [15]. Every minute, more than 350 000 tweets are
posted to Twitter, approximately 3 million Facebook
posts are shared, 4 million Google queries are submitted,
and over 170 million e-mails are sent [16,17]. In more
human terms, the average office worker sees over 120 e-
mails per day [18], the typical teen in the United States
sends over 60 text messages per day from their mobile
phones [19] and the average Facebook user writes 25 com-
ments daily [20]. In short, the amount of language data
generated by humans on a minute-by-minute basis
around the world is nothing short of staggering.
As with the unprecedented availability of human-gener-
ated data, the field of psychology has witnessed a recent
cascade in psychometric techniques that are well-suited
to a big data research culture. Of the more recent psy-
chological assessment methods, perhaps the most acces-
sible and refined to date is that of automated language
analysis, which is currently experiencing rapid adoption
and growth across a wide range of academic fields. His-
torically, psychologists have long believed that a person’s
words can be revealing of deeper, meaningful psycholog-
ical constructs [21 23]. For example, classical research on
motivation found that the individual’s personal strivings,
such as the needs for affiliation and achievement, were
manifest in their everyday words [24], and it has long been
believed that linguistic cues can be used to identify
different states of consciousness [25]. However, the mod-
ern rejuvenation of language research in the field of
Current Opinion in Behavioral Sciences 2017, 18:63 68
personality psychology has been primarily driven by
the adoption of modern statistical methods and techno-
logical innovations, such as the boom of personal com-
puting power and data accessibility [26].
Unlike most classical research on language and psychol-
ogy, which typically treated linguistic measures as indi-
cators of a person’s transient mental state [14,27], several
key studies were conducted early on in the current
language analysis renaissance which demonstrated that
the properties of language-based psychological measures
behave in much the same way as traditional measures of
personality. For example, Pennebaker and King [28]
explored the psychometric properties of language as a
psychological measure, finding that the majority of mea-
sures provided by the Linguistic Inquiry and Word Count
method [29] exhibited all of the hallmarks of a standard
individual differences measure: test retest reliability,
external validity, and internal consistency. A considerable
amount research within the LIWC domain has expanded
these initial findings, establishing the word-counting par-
adigm as a robust tool for measuring stable individual
differences [30,31,32].
In the modern research world, where psychologically-
relevant data is available in great abundance, psychomet-
ric techniques like language analysis allows researchers to
indirectly probe and better understand how lower level
psychological processes function and interact to manifest
in the form of personality in the real world. In other words,
techniques such as language analysis are particularly well-
suited to the proximal measurement the lower level
processes that cohere to form personality, especially in
relation to traditional self-report measures. Countless
patterns of attention, behaviors, and emotions are deeply
embedded in a person’s language [31], and psychologists
now have access to an ever-growing number of methods to
extract these patterns for deeper study.
Given the modern surge of language data, as well as
methods for extracting psychological information from
such data, a logical next step for social scientists is to
begin benefiting from the trait-like qualities of language-
based measures in psychological research. In the current
climate of the ‘big data’ revolution, many of the logistical
properties for which self-report measures are often lauded
ring even truer for language-based measures of personal-
ity. While self-reports are relatively easy to collect com-
pared to other measures such as physiological data, lan-
guage analysis often relies on data that already
exists. Moreover, pre-existing digital data from the web,
smart phones, and social media are inherently ecologically
valid, having originated from thoughts and behaviors that
occur in the absence of researcher intervention.
It is vital to note that the analysis of language for person-
ality research can be performed at scale in nearly any
www.sciencedirect.com

context where language data exists, bypassing the need to
recruit and collect constrained self-report measures.
While it is a harrowing and costly task to collect self-
reported neuroticism from thousands of people, neuroti-
cism’s underlying processes can be measured in millions
of Reddit users’ language in an afternoon. As the number
of people who use digital technology continues to
increase around the world, along with the trails of psy-
chologically actionable data that are left behind, it is
imperative that new methods be adopted that are able
to make good use of this data by capitalizing on the
growing technological infrastructure (e.g. text messages,
institutional databases, and social media). In failing to
adapt to the new big data world, many personality
researchers will be resigned solely to the study self-
theories, and only in samples that are directly accessible
and motivated to fill out questionnaires.
The language-based measurement of
personality
In contrast to most lexical theories of personality, which
posit that descriptions of important personality traits
are embedded within language in general [33 35], it is
implicit to current psychological language analysis
research that several characteristics of someone’s person-
ality are embedded in their unique patterns of language
use. However, both approaches generally assume a taxo-
nomical structure of personality that is, personality as a
broad, abstract construct is composed of lower-level psy-
chological processes and behavioral tendencies [36].
The taxonomical structure of personality, both within a
general personality psychology framework as well as
within a language-based personality framework, is central
to performing meaningful personality psychology re-
search. For example, the underlying components of extra-
version have been well-established to date across various
methodologies: relative to introverts, extraverts generally
engage in more social activity [37], experience greater
positive affect and well-being [38], and are reactive to
external stimulation [39 41]. Indeed, language-based
personality research consistently and successfully finds
the same basic underpinning processes of extraversion.
Relative to their introverted counterparts, extraverts tend
to use higher rates of social words, words indicative of
positive emotions, and language that is representative of
an external focused (i.e. fewer 1st person singular pro-
nouns) [42].
The two dominant modes of language–
personality research
Predicting self-report measures
Contemporary language analysis research typically adopts
1 of 2 overarching approaches. In the first approach,
researchers seek to build language-based models of per-
sonality that approximate the data found in ubiquitous
self-report based studies. In simple terms, one of the
www.sciencedirect.com
Language-based personality Boyd and Pennebaker 65
most common approaches to language personality
research involves using linguistic measures to estimate
how people fill out personality self-report questionnaires.
For example, Yarkoni [43] explored LIWC-based and
word-based statistical models of personality in bloggers’
texts to predict their self-reported Big 5 scores (both
overall scores as well as facet-level measures). Similarly,
Schwartz et al. [44] adopted an ‘open-vocabulary’
approach to predicting Big 5 self-report measures from
Facebook status updates. Such an approach is currently
the dominant paradigm in language personality research
and is primarily driven by research teams that lean
heavily on a predictive modeling background, crossing
boundaries from information sciences to social sciences
[45,46,47,48,49].
Under the ‘estimate self-reports using language’ model of
study, researchers are ultimately seeking to maximize
their account of variance in questionnaire scores via
lexical features, and their studies often yield impressive
results. Nevertheless, it is conceptually problematic to
treat personality as measured by self-report question-
naires as ‘ground truth’ scores for personality research.
In part, well-established limitations of such measures,
such as self-knowledge constraints and response biases
[50], restrict these language-based models of personality
to self-theories. More important is that aggregate mea-
sures of personality are distal abstractions of the very
behaviors, feelings, and thoughts that we seek to under-
stand. In estimating peoples’ self-reported neuroticism
from language, for example, questionnaire scores are
treated as a ‘real’ thing that can be objectively measured
rather than a collection of supporting psychological pro-
cesses. In other words, this paradigm treats self-reported
personality as a ‘gold standard’ while failing to acknowl-
edge the flaws that they acquire as a part of the oper-
ationalization and data collection process.
Measuring personality processes
It is more consistent with modern theories of personality,
then, when the use of language in personality research
adopts a relatively more atomic demeanor to measuring
personality processes, rather than predicting traits as a
generalized whole. This alternative approach to lan-
guage-based research in psychology, while not new, has
begun to see increasing adoption among researchers in
social and personality psychology.
Recent research has found that many basic psychological
tendencies that give rise to broader individual differences
are deeply embedded in language use. For example,
linguistic measures of various cognitive patterns are par-
ticularly predictive of objective outcomes such as college
grades [51,52], life expectancy [53,54], and resilience to
trauma [55,56]. Moreover, language-based measures of
personality processes have reliable, trait-like properties
[28,30]. Further still, such measures are often more
Current Opinion in Behavioral Sciences 2017, 18:63 68
66 Big data in the behavioral sciences
predictive of specific, concrete behaviors than traditional
self-report measures, providing both stronger and broader
predictive coverage [57,58]. Finally, such low-level
measures of personality processes may still be aggregated
into higher-level abstractions for generalized predictive
purposes, much like the work of Yarkoni [43], Schultheiss
[59], Schwartz et al. [44], and others.
Particularly vital to personality psychology as a field,
language-based measures of personality processes allow
researchers to better understand the psychological features
that underpin personality, thereby addressing classical
criticisms of trait research being primarily descriptive
rather than explanatory [60,61]. For example, Carey
et al. [62] extensively debunked the widespread mis-
conception that narcissists are prone to disproportionate
self-focus by measuring rates of 1st person singular pro-
noun use. In their work, the researchers noted that other
psychological processes related to a broader social orien-
tations, including interaction style (e.g. disagreeable
social behaviors) and disinhibition (e.g. impulsivity and
sensation seeking), are more central pillars of the narcis-
sistic personality [63,64]. Similarly, basic motivational
processes that underpin traits such as political ideology,
mindfulness, values, social personality, and motivation
have been identified and integrated into theoretical
understandings of the constructs [13,65 69] some-
thing that is simply not possible with an approach that
relies purely on self-report estimation.
Conclusions
While we have known for some time that self-report
questionnaires suffer from critical limitations, personality
psychologists have been slow to adopt alternatives. As
personality and social psychology have become increas-
ingly integrated [70], research from labs all over the world
have found that a person’s words say more than what
meets the eyes (or ears). Thousands of published studies
have demonstrated that language, a powerfully social
component of human behavior, contains deeply embed-
ded and hidden information about not just social pro-
cesses, but also psychological functioning, attentional
processes, behaviors, and other important psychological
constructs that are absolutely paramount to our under-
standing of personality. Moreover, new methods of quan-
tifying psychological processes from language are
constantly being created. The abundance of language-
based methods designed to improve our understanding of
psychological processes are particularly relevant and
applicable to the modern digital age, where human-
generated data is created a rate far beyond what we
can currently process.
The future of personality research will continue to inno-
vate with new methods to capture the psychological
processes that are embedded in the massive digital trail
of human data. Language analysis for personality research
Current Opinion in Behavioral Sciences 2017, 18:63 68
is a low-hanging fruit that is ripe for the picking. In the
coming years, the integration of objective, multimodal
data such as quantified images, language, audio, mobile
sensor data, and internet behaviors into more refined
measures of personality and its supporting psychological
processes are likely to occur. Given that the road has
already begun to be paved in words, however, there has
never been a better time to transition away from self-
reports and toward language analysis as a foundational
method in personality research.
Conflict of interest statement
Nothing declared.
References and recommended reading
Papers of particular interest, published within the period of review,
have been highlighted as:
 of special interest
 of outstanding interest
1. Kagan J: Galen’s Prophecy: Temperament in Human Nature. Basic
Books; 1998.
2. Freud S: Three Essays on the Theory of Sexuality. Basic Books;
1905.
3. Thomas A, Chess S, Birch HG: The origin of personality. Sci Am
1970, 223:102 109.
4. Allport GW: Pattern and Growth in Personality. Holt, Reinhart &
Winston; 1961.
5. Allport FH, Allport GW: Personality traits: their classification
and measurement. J Abnorm Soc Psychol 1921, 16:6 40.
6. Cattell RB: The Scientific Analysis of Personality. Penguin Books;
1965.
7. Eysenck HJ: The scientific study of personality. Br J Stat
Psychol 1953, 6:44 52.
8. Costa PT, McCrae RR: Revised NEO Personality Inventory
(NEO PI R) and NEO Five Factor Inventory (NEO FFI) Professional
Manual. 1992.
9. Goldberg LR: The structure of phenotypic personality traits.
Am Psychol 1993, 48:26 34.
10. McCrae RR, Costa PT: Self concept and the stability of
personality: cross sectional comparisons of self reports and
ratings. J Pers Soc Psychol 1982, 43:1282 1292.
11. Morgeson FP, Campion MA, Dipboye RL, Hollenbeck JR,
Murphy K, Schmitt N: Reconsidering the use of personality
tests in personnel selection contexts. Pers Psychol 2007,
60:683 729.
12. Rhodes RE, Smith NEI: Personality correlates of physical
activity: a review and meta analysis. Br J Sports Med 2006,
40:958 965.
13. Wojcik SP, Hovasapian A, Graham J, Motyl M, Ditto PH:
 Conservatives report, but liberals display, greater happiness.
Science 2015, 347:1243 1246.
A long running debate in psychology is whether conservatives or liberals
are more happy, in general. While past research has repeatedly found that
conservatives report greater happiness in self report paradigms, liberals
actually exhibit greater happiness, as quantified in their language and
other behavioral measures.
14. Stiles WB: Describing Talk: A Taxonomy of Verbal Response
Modes. 1992.
15. International Telecommunications Union: ICT Facts and
Figures 2016. 2016.
www.sciencedirect.com

16. Micro Focus. How Much Data is Created on the Internet Each
Day [Internet]. [September 8, 2016], [no volume].
17. Internet Live Stats: Twitter Usage Statistics [Internet]. [date
unknown], [no volume].
18. Radicati Group: Email Statistics Report, 2015 2019.. 2015.
19. Pew Internet Project: Teens, Smartphones, and Texting. 2012.
20. Leonard H: This is what an average user does on Facebook
[Internet]. 2013. [no volume].
21. Freud S: On Aphasia. International Universities Press; 1891.
22. Lasswell HD, Lerner D, de Sola Pool I: The Comparative Study of
Symbols: An Introduction. Stanford University Press; 1952.
23. Stone PJ, Dunphy DC, Smith MS, Ogilvie DM: The General
Inquirer: A Computer Approach to Content Analysis. M.I.T. Press;
1966.
24. McClelland DC, Atkinson JW, Clark RA, Lowell EL: The
Achievement Motive. Irvington; 1953.
25. Martindale C: The grammar of altered states of consciousness:
a semiotic reinterpretation of aspects of psychoanalytic
theory. Psychoanal Contemp Thought 1975, 4:331 354.
26. Boyd RL: Psychological text analysis in the digital humanities.
In Data Analytics in the Digital Humanities. Edited by Hai Jew S.
Springer International Publishing; 2017:161 189.
27. Gottschalk LA: The unobtrusive measurement of psychological
states and traits. In Text Analysis for the Social Sciences:
Methods for Drawing Statistical Inferences from Texts and
Transcripts. Edited by Roberts CW. Erlbaum; 1997:117 129.
28. Pennebaker JW, King LA: Linguistic styles: language use as an
individual difference. J Pers Soc Psychol 1999, 77:1296 1312.
29. Pennebaker JW, Francis ME: Linguistic Inquiry and Word Count
(LIWC): A Computer based Text Analysis Program. 1999.
30. Boyd RL, Pennebaker JW: Did Shakespeare Write Double
 Falsehood? Identifying Individuals by Creating Psychological
Signatures with Text Analysis. Psychol Sci 2015, 26:570 582.
The authors used language based measures of personality processes to
successfully differentiate multiple people, ultimately determining that
Shakespeare was the primary author of a disputed play. By using
psychological language analysis, the differentiating linguistic measures
were able to be interpreted in light of observer reports of different people,
providing high convergence.
31. Tausczik YR, Pennebaker JW: The Psychological Meaning of
Words: LIWC and Computerized Text Analysis Methods.
J Lang Soc Psychol 2010, 29:24 54.
32. Pennebaker JW: The Secret Life of Pronouns: What Our Words Say
About Us. 2011 http://dx.doi.org/10.1093/llc/fqt006.
33. Hogan R: A socioanalytic perspective on the five factor model.
In The Five factor Model of Personality: Theoretical Perspectives.
Edited by Wiggins JS. Guilford; 1996:163 179.
34. John OP, Angleitner A, Ostendorf F: The lexical approach to
personality: a historical review of trait taxonomic research. Eur
J Pers 1988, 2:171 203.
35. Golubkov SV: The language personality theory: an integrative
approach to personality on the basis of its language
phenomenology. Soc Behav Pers 2002, 30:571 578.
36. Costa PT, McCrae RR: Domains and facets: hierarchical
personality assessment using the revised NEO personality
inventory. J Pers Assess 1995, 64:21 50.
37. Snyder M: The influence of individuals on situations:
implications for understanding the links between personality
and social behavior. J Pers 1983, 51:497 516.
38. Furnham A, Brewin CR: Personality and happiness. Pers Individ
Dif 1990, 11:1093 1096.
39. Campbell JB: Differential relationships of extraversion,
impulsivity, and sociability to study habits. J Res Pers 1983,
17:308 314.
www.sciencedirect.com
Language-based personality Boyd and Pennebaker 67
40. Philipp RL, Wilde GJS: Stimulation seeking behaviour and
extraversion. Acta Psychol (Amst) 1970, 32:269 280.
41. Smits DJM, Boeck PD: From BIS/BAS to the big five. Eur J Pers
2006, 20:255 270.
42. Mairesse F, Walker MA, Mehl MR, Moore RK: Using linguistic
cues for the automatic recognition of personality in
conversation and text. J Artif Intell Res 2007, 30:457 500.
43. Yarkoni T: Personality in 100,000 words: a large scale analysis
of personality and word use among bloggers. J Res Pers 2010,
44:363 373.
44. Schwartz HA, Eichstaedt JC, Kern ML, Dziurzynski L, Ramones SM,
Agrawal M, Shah A, Kosinski M, Stillwell D, Seligman MEP et al.:
Personality, gender, and age in the language of social media:
the open vocabulary approach. PLoS ONE 2013, 8:1 16.
45. Badenes H, Bengualid MN, Chen J, Gou L, Haber E, Mahmud J,
Nichols JW, Pal A, Schoudt J, Smith BA et al.: System U:
automatically deriving personality traits from social media for
people recommendation. In Proceedings of the 8th ACM
Conference on Recommender Systems; ACM: 2014:373 374.
46. Chen J, Haber E, Kang R, Hsieh G, Mahmud J: Making use of
 derived personality: the case of social media ad targeting. In
Proceedings of the Ninth International AAAI Conference on Web
and Social Media. 2015:51 60.
The authors used language samples to estimate self report scores for the
Big 5, then used these estimated scores to model responsiveness to
targeted advertising. This work is an example of the many ways in which
personality is often misconceptualized when studied from a predictive
modeling viewpoint.
47. Collins S, Sun Y, Kosinski M, Stillwell D, Markuzon N: Are you
satisfied with life? Predicting satisfaction with life from
Facebook. In In Social Computing, Behavioral Cultural Modeling,
and Prediction: 8th International Conference, SBP 2015,
Washington, DC, USA, March 31 April 3, 2015, Proceedings.
Edited by Agarwal N, Xu K, Osgood N. Social Computing,
Behavioral Cultural Modeling, and Prediction: 8th International
Conference, SBP 2015, Washington, DC, USA, March 31 April 3,
2015, ProceedingsSpringer International Publishing: 2015:24 33.
48. Komisin M, Guinn C: Identifying personality types using
document classification methods. In Proceedings of the
Twenty Fifth International Florida Artificial Intelligence Research
Society Conference. 2012:232 237.
49. Park G, Schwartz HA, Eichstaedt JC, Kern ML, Kosinski M,
 Stillwell DJ, Ungar LH, Seligman MEP: Automatic personality
assessment through social media language. J Pers Soc
Psychol 2015, 108:934 952.
One of several impressive studies where the stated goal is to maximize
the variance accounted for in self report personality questionnaires. The
authors of this study demonstrated a new approach to estimating how
people typically respond to self reported measures of personality by
using the language that people share on social media.
50. Paulhus DL, Vazire S: The self report method. Handbook of
Research Methods in Personality Psychology; Guilford: 2007:224
239.
51. Pennebaker JW, Chung CK, Frazee J, Lavergne GM, Beaver DI:
When small words foretell academic success: the case of
college admissions essays. PLoS ONE 2015, 9:e115844.
52. Robinson RL, Navea R, Ickes W: Predicting final course
performance from students’ written self introductions. J Lang
Soc Psychol 2013, 32:469 479.
53. Pressman SD, Cohen S: Use of social words in autobiographies
and longevity. Psychosom Med 2007, 69:262 269.
54. Penzel IB, Persich MR, Boyd RL, Robinson MD: Linguistic
evidence for the failure mindset as a predictor of life span
longevity. Ann Behav Med 2016 http://dx.doi.org/10.1007/
s12160 016 9857 x.
55. Pennebaker JW, Mayne TJ, Francis ME: Linguistic predictors of
adaptive bereavement. J Pers Soc Psychol 1997, 72:863 871.
56. D’Andrea W, Chiu PH, Casas BR, Deldin P: Linguistic predictors
of post traumatic stress disorder symptoms following
11 September 2001. Appl Cogn Psychol 2012, 26:316 323.
Current Opinion in Behavioral Sciences 2017, 18:63 68
 68 Big data in the behavioral sciences
57. Boyd R, Wilson S, Pennebaker J, Kosinski M, Stillwell D, Mihalcea R:
 Values in words: using language to evaluate and understand
personal values. In Proceedings of the Ninth International AAAI
Conference on Web and Social Media. 2015:31 40.
The authors introduced a new method for establishing language based
measures of core values. This research found that the language based
measures of values showed poor convergence with self reported values
yet were vastly superior in terms of predictive strength and coverage
when modeling the important relationship between values and behavior
found in the real world.
58. Fast LA, Funder DC: Personality as manifest in word use:
correlations with self report, acquaintance report, and
behavior. J Pers Soc Psychol 2008, 94:334 346.
59. Schultheiss O: Are implicit motives revealed in mere words?
Testing the marker word hypothesis with computer based
text analysis. Front Psychol 2013, 4:748.
60. Pervin LA: A critical analysis of current trait. Psychol Inq 1994,
5:103 113.
61. Wilson M, De Boeck P: Descriptive and explanatory item
response models. In Explanatory Item Response Models: A
Generalized Linear and Nonlinear Approach. Edited by Wilson M,
De Boeck P. Springer; 2004:43 74.
62. Carey AL, Brucks MS, Küfner ACP, Holtzman NS, große Deters F,
 Back MD, Donnellan MB, Pennebaker JW, Mehl MR: Narcissism
and the use of personal pronouns revisited. J Pers Soc Psychol
2015, 109:e1 e15.
The authors found that, contrary to both layperson and expert assump
tions, narcissism is not associated with more self focused language. This
research is a prime example of how psychological language analysis can
be extremely informative for personality theory and clarifying misguided
assumptions.
Current Opinion in Behavioral Sciences 2017, 18:63 68
View publication stats
63. Vazire S, Funder DC: Impulsivity and the self defeating
behavior of narcissists. Personal Soc Psychol Rev 2006, 10:154
165.
64. Miller JD, Campbell WK, Young DL, Lakey CE, Reidy DE,
Zeichner A, Goodie AS: Examining the relations among
narcissism, impulsivity, and self defeating behaviors. J Pers
2009, 77:761 794.
65. Collins SE, Chawla N, Hsu SH, Grow J, Otto JM, Marlatt GA:
Language based measures of mindfulness: Initial validity and
clinical utility. Psychol Addict Behav 2009, 23:743 749.
66. Fetterman AK, Boyd RL, Robinson MD: Power versus affiliation
in political ideology. Personal Soc Psychol Bull 2015, 41:1195
1206.
67. Johnsen J AK, Vambheim SM, Wynn R, Wangberg SC: Language
of motivation and emotion in an internet support group for
smoking cessation: explorative use of automated content
analysis to measure regulatory focus. Psychol Res Behav
Manag 2014, 7:19 29.
68. Kacewicz E, Pennebaker JW, Davis M, Jeon M, Graesser AC:
Pronoun use reflects standings in social hierarchies. J Lang
Soc Psychol 2013, 33:125 143.
69. Reysen S, Pierce L, Mazambani G, Mohebpour I, Puryear C,
Snider JS, Gibson S, Blake ME: Construction and initial
validation of a dictionary for global citizen linguistic markers.
Int J Cyber Behav Psychol Learn 2014, 4:1 15.
70. Snyder M, Deaux K: Personality and social psychology:
crossing boundaries and integrating perspectives. In The
Oxford Handbook of Personality and Social Psychology. Edited by
Snyder M, Deaux K. Oxford University Press; 2012:3 9.
www.sciencedirect.com